Filters








28,210 Hits in 6.8 sec

Better Process Mapping and Sparse Quadratic Assignment [article]

Christian Schulz and Jesper Larsson Träff and Konrad von Kirchbach
2019 arXiv   pre-print
We address the problem as a quadratic assignment problem (QAP), and present algorithms to construct initial mappings of processes to processors, and fast local search algorithms to further improve the  ...  Communication and topology aware process mapping is a powerful approach to reduce communication time in parallel applications with known communication patterns on large, distributed memory systems.  ...  [5] and others, and model the embedding problem as a quadratic assignment problem (QAP): Find a one-to-one mapping Π of processes to PEs which minimizes the overall communication cost.  ... 
arXiv:1702.04164v2 fatcat:hgid7p4fjjbzbbzcdlwntp2kjm

Better Process Mapping and Sparse Quadratic Assignment *

S Costas, Iliopoulos, P Solon, Simon Pissis, Rajeev Puglisi, Raman
2017 15 Leibniz International Proceedings in Informatics Schloss Dagstuhl-Leibniz-Zentrum für Informatik   unpublished
We address the problem as a quadratic assignment problem (QAP), and present algorithms to construct initial mappings of processes to processors as well as fast local search algorithms to further improve  ...  Such a mapping can be computed by solving a corresponding quadratic assignment problem (QAP) which is a hard optimization problem.  ...  S E A 2 0 1 7 4:8 Better Process Mapping and Sparse Quadratic Assignment algorithms.  ... 
fatcat:bkovyzqmefb6jpezajm64puibu

On the Sparseness and Generalization Capability of Least Squares Support Vector Machines

Aijun Yan, Xiaoqian Huang, Hongshan Shao
2015 Journal of Systems Science and Information  
Then, pick the subset that has the largest quadratic entropy to train and prune, and repeat this process until the cumulative error rate reaches the condition requirement.  ...  AbstractCompared with standard support vector machines (SVM), sparseness is lost in the modeling process of least squares support vector machines (LS-SVM), causing limited generalization capability.  ...  pointed out that a model with high sparseness tends to have better generalization performance and computing speed in machine learning.  ... 
doi:10.1515/jssi-2015-0279 fatcat:sifve2uiebcmphgc4ourjgskwq

Digitizing the coral reef: machine learning of underwater spectral images enables dense taxonomic mapping of benthic habitats [article]

Daniel Schürholz, Arjun Chennu
2022 bioRxiv   pre-print
With low annotation effort (2% pixels) and no external data, this workflow enables accurate (Fbeta 87%) survey-scale mapping, with unprecedented thematic and spatial detail.  ...  All 500+ million pixels were assigned to 43 labels at taxonomic family, genus or species level for corals, algae, sponges, or to substrate labels such as sediment, turf algae and cyanobacterial mats.  ...  Acknowledgments We would like to thank Carsten John and Oliver Artmann at the Max Planck Institute for Marine Microbiology for their IT support.  ... 
doi:10.1101/2022.03.28.485758 fatcat:s2ltvjgjvrbvhjhu2iogwzsicm

Efficient and Robust Shape Correspondence via Sparsity-Enforced Quadratic Assignment [article]

Rui Xiang, Rongjie Lai, Hongkai Zhao
2020 arXiv   pre-print
These two ingredients can improve and increase the number of anchor points quickly while reducing the computation cost in each quadratic assignment iteration significantly.  ...  In this work, we introduce a novel local pairwise descriptor and then develop a simple, effective iterative method to solve the resulting quadratic assignment through sparsity control for shape correspondence  ...  We introduce our quadratic assignment model based on a local pairwise descriptor in Section 2 and then present an efficient iterative algorithm to solve the quadratic assignment problem with sparsity control  ... 
arXiv:2003.08680v2 fatcat:26onaidqmbdpfhtovw53lyrdpi

Adding Depth to Cartoons Using Sparse Depth (In)equalities

D. Sýkora, D. Sedlacek, S. Jinchao, J. Dingliana, S. Collins
2010 Computer graphics forum (Print)  
In comparison to previous depth assignment techniques our solution requires minimal user effort and enables creation of consistent pop-ups in a matter of seconds.  ...  Its key advantage is that it completely avoids inputs requiring knowledge of absolute depth and instead uses a set of sparse depth (in)equalities that are much easier to specify.  ...  Kavan and anonymous reviewers for their fruitful comments. Images used in this paper are courtesy of studios UPP, DMP & Anifilm, P. Koutský, and K. Mlynaříková.  ... 
doi:10.1111/j.1467-8659.2009.01631.x fatcat:jpe7ra3erjdljbis63yeodu6c4

Efficient and Robust Shape Correspondence via Sparsity-Enforced Quadratic Assignment

Rui Xiang, Rongjie Lai, Hongkai Zhao
2020 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)  
These two ingredients can improve and increase the number of anchor points quickly while reducing the computation cost in each quadratic assignment iteration significantly.  ...  In this work, we introduce a novel local pairwise descriptor and then develop a simple, effective iterative method to solve the resulting quadratic assignment through sparsity control for shape correspondence  ...  We introduce our quadratic assignment model based on a local pairwise descriptor in Section 2 and then present an efficient iterative algorithm to solve the quadratic assignment problem with sparsity control  ... 
doi:10.1109/cvpr42600.2020.00953 dblp:conf/cvpr/XiangLZ20 fatcat:d6iexpserrg7vovmerwsci7ora

Depth map coding using graph based transform and transform domain sparsification

Gene Cheung, Woo-Shik Kim, Antonio Ortega, Junichi Ishida, Akira Kubota
2011 2011 IEEE 13th International Workshop on Multimedia Signal Processing  
For blocks without detected prominent edges, the synthesized view's distortion sensitivity to depth map errors is low, and TDS can effectively identify a sparse depth signal in fixed DCT domain within  ...  For blocks with detected prominent edges, the synthesized view's distortion sensitivity to depth map errors is high, and the search space of good depth signals for TDS to find sparse representations in  ...  The view synthesis is performed by warping the left and right views to the target view position (middle view), then blending process is applied if more than one pixels are mapped to the same position,  ... 
doi:10.1109/mmsp.2011.6093810 dblp:conf/mmsp/CheungKOIK11 fatcat:i2wri3we2rezdjmwb5jbeonjn4

Improving IMRT delivery efficiency with reweighted L1-minimization for inverse planning

Hojin Kim, Stephen Becker, Rena Lee, Soonhyouk Lee, Sukyoung Shin, Emmanuel Candès, Lei Xing, Ruijiang Li
2013 Medical Physics (Lancaster)  
Results: The proposed method yields simpler fluence-maps than the quadratic and conventional TV based techniques.  ...  Methods: First-order total-variation (TV) minimization (min.) based on L1-norm has been proposed to reduce the complexity of fluence-map in IMRT by generating sparse fluence-map variations.  ...  Figure 2 FIG. 2 .FIG. 3 . 223 compares three resultant fluence-maps optimized by quadratic, conventional TV min., and the proposed Fluence-maps acquired by quadratic min., TV min. and reweighted TV min  ... 
doi:10.1118/1.4811100 pmid:23822423 pmcid:PMC3702602 fatcat:aqk2dpo73nck3k3c33mdiykb7m

Modelling Lichen Abundance for Woodland Caribou in a Fire-Driven Boreal Landscape

Joseph A. Silva, Scott E. Nielsen, Clayton T. Lamb, Christine Hague, Stan Boutin
2019 Forests  
We interpolated the best lichen presence and lichen abundance models to create spatial layers and combined them to generate a map that provides a reasonable estimation of lichen biomass (R2 = 0.39) for  ...  We encourage researchers and managers to use our method as a basic framework to map the abundance of ground lichens across fire-prone, boreal caribou ranges.  ...  Acknowledgments: We are grateful for the generous and extensive support of Ontario Parks and the Ontario Ministry of Natural Resources and Forestry.  ... 
doi:10.3390/f10110962 fatcat:cm23ouvg4zbrbhxi64v7vbbtki

Forward-Decoding Kernel-Based Phone Recognition

Shantanu Chakrabartty, Gert Cauwenberghs
2002 Neural Information Processing Systems  
Training over very large data sets is accomplished using a sparse probabilistic support vector machine (SVM) model based on quadratic entropy, and an on-line stochastic steepest descent algorithm.  ...  Forward decoding kernel machines (FDKM) combine large-margin classifiers with hidden Markov models (HMM) for maximum a posteriori (MAP) adaptive sequence estimation.  ...  The 'Gini' index provides a lower bound of the dual logistic functional, and its quadratic form produces sparse solutions as with support vector machines.  ... 
dblp:conf/nips/ChakrabarttyC02 fatcat:npi2lwv275fxvpv26ysfaqykhe

Potential Convolution: Embedding Point Clouds into Potential Fields [article]

Dengsheng Chen and Haowen Deng and Jun Li and Duo Li and Yao Duan and Kai Xu
2021 arXiv   pre-print
Recently, various convolutions based on continuous or discrete kernels for point cloud processing have been widely studied, and achieve impressive performance in many applications, such as shape classification  ...  , scene segmentation and so on.  ...  The naive implementation of˜1 ,˜ ,˜2 is inefficient, and we find that the computation process of linear and quadratic functions can be highly paralleled by formulating them as point-wise convolution operations  ... 
arXiv:2104.01754v1 fatcat:qazl5ngjxffx3gfas4tdt3wily

A robust multimedia surveillance system for people counting

Zeyad Q. H. Al-Zaydi, David L. Ndzi, Munirah L. Kamarudin, Ammar Zakaria, Ali Y. M. Shakaff
2016 Multimedia tools and applications  
Each algorithm is designed using a novel combination of pixel-wise, motion-region, grid map, background segmentation using Gaussian mixture model (GMM) and edge detection.  ...  The mean deviation error, mean squared error and the mean absolute error of the proposed system are less than 0.1, 16.5 and 3.1, respectively, for the Mall dataset and less than 0.07, 5.5 and 1.9, respectively  ...  the value 257 so that quadratic programming can be used to find 257 density values instead of 256.Divide the frame into cells (as a grid map) and the number of people in each cell =W+B is the grid map  ... 
doi:10.1007/s11042-016-4156-x fatcat:jor44vyzsjddlpnxajmaokac4i

A network of networks processing model for image regularization

Ling Guan, J.P. Anderson, J.P. Sutton
1997 IEEE Transactions on Neural Networks  
This method is able to provide fast, quality imaging in early vision, and its replicating structure and sparse connectivity readily lend themselves to hardware implementations.  ...  The method is motivated by the fact that natural image formation involves both local processing and globally coordinated parallel processing.  ...  Equations (5) and (6) show the mapping of the quadratic programming model to an ordinary attractor network.  ... 
doi:10.1109/72.554202 pmid:18255621 fatcat:k35smrcxbnbqbob7xq3rqdxrda

Efficient hybrid search for visual reconstruction problems

Shang-Hong Lai, Baba C. Vemuri
1999 Image and Vision Computing  
variables to better constraint the line processes.  ...  The performance of our algorithm is demonstrated via experimental results on the sparse data surface reconstruction and the image restoration problem. ᭧  ...  Acknowledgements Supported in part by the NSF grant ECS-9210648 and the Whitaker Foundation.  ... 
doi:10.1016/s0262-8856(98)00088-2 fatcat:5vk3ijcp7fcklonrfektllctny
« Previous Showing results 1 — 15 out of 28,210 results